instruction

Blog

OpenAI GPT-4.1 models promise improved coding and instruction following

“GPT‑4.1 mini is a significant leap in small model performance, even beating GPT‑4o in many benchmarks. It matches or exceeds GPT‑4o in intelligence evals while reducing latency by nearly half and reducing cost by 83%,” the announcement said. “For tasks that demand low latency, GPT‑4.1 nano is our fastest and cheapest model available. It delivers exceptional performance at a small…